A new sensitive in vitro assay for human thyroid stimulator (HTS) was developed using human thyroid adenoma cells in monolayer culture. After being cultured for 2 days, the cells were incubated in 0.3 ml Hank's solution without 0.8% NaCl (medium 1) and with thyroid stimulator (bovine TSH or 3 mg patient serum immunoglobulin G) at 37 C for 2 h. The cAMP generated in the cells and the medium during the incubation was measured by RIA. The assay was sensitive enough to elicit a 1.7- to 7.9-fold increase in cAMP at a TSH concentration of 10 microU/ml. HTS was detected in 33 (82.5%) of the 40 patients with untreated graves' disease using this assay system. In Hank's solution (medium 2), however, HTS was detected in only 5 (23.8%) of the 21 patients with untreated GRaves' disease. cAMP increment upon stimulation by either TSH or HTS in medium 1 was larger than that in medium 2, and the difference in the response to HTS using the two media was much greater than that in the response to TSH. Therefore, all HTS-immunoglobulin G studies showed higher activity using medium 1 than using medium 2 when expressed as bovine TSH equivalent. Analysis by the Lineweaver-Burk plot of dose-response curves of the effect of TSH and HTS stimulation on cAMP increment showed an increase in the Km upon the addition of NaCl to the medium. A similar inhibitory effect of NaCl (150 mM) was also observed in the assay system of human thyroid adenylate cyclase stimulator using crude plasma membrane fractions. In summary: 1) an assay for HTS measuring cAMP production in cultured thyroid adenoma cells was developed and the assay using low NaCL medium was found to be the most sensitive, and 2) the inhibitory effect of NaCl on the response to HTS was much greater than that on the response to TSH. These data suggest different behaviors of these two stimulators at their receptor sites.